| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
| |
This reverts commit 17777040c63b7bc8586fa77bfa1219c0d96a9dd9.
This version is still too new for gnome-build-meta and flatpaks,
let's try to do without it.
|
|
|
|
|
|
|
|
|
| |
This is a meson 0.50 feature which we don't strictly need to rely on.
According to meson docs: "When omitted it defaults to true when
install_dir is set and not empty, false otherwise."
In all places we use it we (obviously) define install_dir, so the
files should be installed without it being specifically told so.
|
|
|
|
|
|
|
|
|
| |
Ensure the namespace is not part of the includes. This is a kludge and
shouldn't ever happen, except it does (in continuous, induced by vala).
The next big thing that would happen here is a rewrite of libtracker-sparql
vala bits so we don't need this merging anymore, so I won't put much
thought on it.
|
|
|
|
|
| |
Ebassi points out it's actually what g-ir-compile expects, so
let's cross fingers again...
|
|
|
|
|
|
|
| |
This is another change needed for tracker-miners to use the GSettings
schemas from this project when its build as a subproject.
See: https://gitlab.gnome.org/GNOME/tracker-miners/merge_requests/61
|
|
|
|
|
|
|
|
|
| |
Try to help it find the Tracker-2.0.gir target it depends upon.
The command is given a relative path from the build root, so add
it.
Hopefully fixes:
http://build.gnome.org/continuous/buildmaster/builds/2019/05/01/5/build/log-tracker.txt
|
|
|
|
|
|
|
|
| |
These changes are needed so that the 'tracker-miners' project can make
use of the 'tracker' settings schemas when build with the
`tracker=subproject` option. There should be no change to funtionality.
This is related to https://gitlab.gnome.org/GNOME/tracker/issues/24
|
|
|
|
|
|
|
|
|
|
|
| |
We must avoid using meson.build_root() because the value changes if we
are embedded as a subproject in another project.
I discovered this issue while testing
<https://gitlab.gnome.org/GNOME/tracker/merge_requests/74> where it
caused GSettings schemas not to be found when Tracker wasn't installed
into /usr . It didn't cause any problems before, because the test was
able to fall back on the GSettings schemas from /usr.
|
|
|
|
|
|
|
| |
Meson helpfully warns us that this is needed:
WARNING: Project specifies a minimum meson_version '>=0.47' but uses features which were added in newer versions:
* 0.50.0: {'install arg in configure_file'}
|
|
|
|
|
|
| |
We would leak the parsing tree accumulated thus far on errors. Good
news is that it's potentially unlikely, bad news is that it's potentially
big.
|
|\
| |
| |
| |
| |
| |
| | |
tracker-miner: Fix cancellation of g_file_enumerator_next_files_async
Closes #86
See merge request GNOME/tracker!86
|
|/
|
|
|
|
|
|
| |
The async op is not owner of the user data, so it may be actually gone in the
GAsyncReadyCallback. Ensure we only use it on success or on other errors than
cancelled.
Closes: https://gitlab.gnome.org/GNOME/tracker/issues/86
|
|\
| |
| |
| |
| | |
libtracker-sparql: Handle correctly backreferences in TrackerResource tree
See merge request GNOME/tracker!84
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
We do keep a notion of already visited TrackerResources when generating the
output. However that could go wrong if the TrackerResource references itself
or a "parent" TrackerResource.
Add the resources to the done list beforehand, so they are ensured not to be
visited again if such backreferences exist.
Related: https://gitlab.gnome.org/GNOME/tracker-miners/issues/60
|
|/
|
|
|
|
|
|
|
|
| |
Since I got meson 0.50, tracker fails to build with:
undefined reference to `builtin_ontology_resource_data'
when building libtracker-data.so. Do not use the compile_gresource
values individually, this seems to make meson 0.50 happy and should
work with older meson.
|
|\
| |
| |
| |
| | |
docs: Modernize the libtracker-sparql documentation
See merge request GNOME/tracker!81
|
|/
|
|
| |
Still not fixed the Tracker 2.0 references here and there
|
|\
| |
| |
| |
| | |
tests: Remove some functional tests
See merge request GNOME/tracker!72
|
| |
| |
| |
| |
| |
| | |
These tests appear to be either basic smoke tests, or performance tests
that are intended to be timed and executed manually. Some also dump huge
amounts of text to the test log file.
|
|\ \
| | |
| | |
| | |
| | | |
docs: Modernize the libtracker-miner documentation
See merge request GNOME/tracker!78
|
| | | |
|
| | | |
|
|\ \ \
| |/ /
|/| |
| | |
| | |
| | |
| | | |
Fix parsing of doubles with exponential part
Closes #84
See merge request GNOME/tracker!77
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Otherwise there are greediness issues as the definition of DECIMAL
fits within DOUBLE and wins over it, leaving the exponential part
out of the match.
Closes: https://gitlab.gnome.org/GNOME/tracker/issues/84
|
|/ /
| |
| |
| |
| | |
A c&p typo made the DOUBLE_POSITIVE terminal defined upon itself, it
should be defined on top of DOUBLE.
|
|\ \
| | |
| | |
| | |
| | | |
docs: Modernize the libtracker-control documentation
See merge request GNOME/tracker!76
|
|/ / |
|
|\ \
| | |
| | |
| | |
| | | |
Switch to README.md
See merge request GNOME/tracker!75
|
|/ / |
|
| | |
|
| | |
|
| | |
|
| |
| |
| |
| |
| |
| | |
This broke in commit bb88e1d39, since it is valid for the cursor
to leave the GValue unset on NULLs in the requested column. We
should just forward the NULL as we used to.
|
|\ \
| | |
| | |
| | |
| | |
| | |
| | | |
Fix CHANGES_DONE_HINT handling on inotify
Closes tracker-miners#36
See merge request GNOME/tracker!73
|
| | |
| | |
| | |
| | |
| | |
| | | |
I don't know if this is possible, but in case that happens we should
drop the event either way, and neutralize CREATED+DELETED events early
on.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
We cache both CREATED and CHANGED events here, if we get both while
waiting for the CHANGES_DONE_HINT, it makes things more consistent
to cache the first CREATED event.
This did not result in bugs as the miner reacts the same either way.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
use_changed_event refers to FAM and the inability of the GIO monitor
implementation to send CHANGES_DONE_HINT for it. This means we have
to forward CREATED/CHANGED events immediately if we have to rely on
FAM-backed monitors.
However the condition handling this was inverted, which meant we
sucked with file monitors that honored CHANGES_DONE_HINT, and
completely broke updates on FAM monitors.
Closes: https://gitlab.gnome.org/GNOME/tracker-miners/issues/36
|
| | |
| | |
| | |
| | |
| | |
| | | |
In the comment above the "may" verb is key. If the monitor does a
good job at coalescing events that would not be the case, we should
admit the flag being either set or unset here.
|
|\ \ \
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
meson: Enforce build order using generated headers directly
Closes #78
See merge request GNOME/tracker!65
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Both libtracker-fts and libtracker-data end up requiring includes from
tracker-sparql.h, but we have to generate tracker-generated.h first. To
solve this, commit a452173ed introduced a dependency to an intermediate
target (tracker_sparql_intermediate_dep). Use directly the custom target
tracker_sparql_generated_header to enforce build order.
Closes: https://gitlab.gnome.org/GNOME/tracker/issues/78
|
|\ \ \ \
| | | | |
| | | | |
| | | | |
| | | | | |
ci: Upload test suite logs when tests fail
See merge request GNOME/tracker!71
|
| | |_|/
| |/| |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
We sometimes see unreproducible test failures in CI. These are often
hard to debug because they don't occur all the time, and may even be
triggered by something that only happens on a specific CI runner.
As a step towards debugging these, we should upload the test log as
a GitLab artifact. Meson only shows the last 100 lines of the test log
on stdout, and there seems to be no way to cause it to emit more info,
so this should help with debugging
https://gitlab.gnome.org/GNOME/tracker/issues/79.
|
|\ \ \ \
| |/ / /
|/| | |
| | | |
| | | | |
Survive missing ontologies.gvdb file
See merge request GNOME/tracker!68
|
|/ / /
| | |
| | |
| | |
| | | |
We shouldn't unref the manager->ontologies object if loading the
new ontologies failed. Fixes a crash in the fallback code path.
|
| |/
|/| |
|
|\ \ |
|
| | |
| | |
| | |
| | | |
This is unused now.
|
| | |
| | |
| | |
| | |
| | | |
Theoretically at least, since we don't have file monitors on the directory
in that case. The code is now correct at least.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Files that are non-indexable by themselves may trigger is_parent_indexable()
changes. Defer the early return after that.
Closes: https://gitlab.gnome.org/GNOME/tracker/issues/47
|
|/ /
| |
| |
| |
| |
| |
| | |
Leave all checks on TrackerIndexingTree to the TrackerFileNotifier,
and report all events to it. Those might trigger other actions
(eg. policy changes through the TrackerIndexingTree) that shouldn't
go ignored.
|